**6. Cytology-Driven MoA Profiling**

In the last few years, we have witnessed an increase in the number of scientific reports on new potential drug candidates to treat leishmaniases and trypanosomiases. However, the vast majority lack insights or detailed mechanism of action evidence supporting further drug development and clinical trials. In this scenario, cell-based assays offer the contextualized relevance and complexity of living cells to track drug discovery approaches, especially when considering unicellular parasites. Kinetoplastids are classified in this category due to the presence of a kinetoplast—a dense structure made by DNA (kDNA) within their unique mitochondria. Therefore, mitochondrial function monitoring can be applied in order to provide hints on the MoA of drug candidates in the drug discovery pipeline. Cellular bioenergetics analysis based on extracellular flux can phenotypically characterize mitochondrial function and define the energetic status of aerobic and glycolytic metabolism, defining a range from quiescent to energetic profiling [111,112]. This approach was used to monitor oxygen consumption (mitochondrial respiration) vs. medium acidification rate (glycolysis) in *L. infantum* to metabolically characterize SbR mutants and evaluate the oxidative role of gold(I) complexes as metallodrug candidates to treat leishmaniasis [44]. This approach was also considered using host cells experimentally infected with *T. cruzi* intracellular amastigotes, monitoring not only the parasite's metabolism, but mimicking the natural conditions considering the context of endogenous conditions of infected cells [113]. These assays were performed on a Seahorse Extracellular Flux Analyzer, XF series (Agilent), and were initially used to monitor basal mitochondrial metabolism in *T. cruzi*, which is useful for drug screening purposes [114–117]. Microscopic imaging using cell-permeant mitochondrion-selective dyes such as MitoTracker or cell permeant acidotropic fluorophores like LysoTracker can be used to highlight ultrastructural alterations in essential organelles to make inferences about drug action and target elucidation by functional approaches [118]. These dyes can be used in high-content analysis approaches that have been shown as an alternative to monitor not only anti-parasitic drug action but also concomitant host toxicity analysis in the same assay for drug screening purposes [119]. Despite the above-mentioned fluorescent gene reporters, kDNA can be labelled to monitor cell replication for indirect drug activity measurement. The terminal deoxynucleotidyl transferase dUTP nick end labelling (TUNEL) technique allows the specific tagging of blunt DNA ends—a common feature in programmed cell death in mammalian cells. Conventional programmed cell death is not biochemically the same in trypanosomatids, and TUNEL signals are undetectable in trypanosome nuclei (genomic DNA). However, 25% of control (wild type, untreated) cells were reported to have TUNEL-positive kDNA. Treatments with eflornithine, nifurtimox, or melarsoprol did not change TUNEL signal, but pentamidine or suramin exposition reduced it, as an evidence of loss of kDNA following the latter treatments in a cell-cycle-dependent manner [120,121]. Trypanosomatids present closed mitosis (chromosomal condensation and segregation is maintained inside the nucleus during division), and the segregation of their single mitochondrial genome (kinetoplast) can be easily monitored by fluorescent microscopy during cell division in the presence of 4',6'-diamidino-2-phenylindole (DAPI, a DNA-intercalating dye. This feature can be tracked under drug treatment to make inferences about mitosis or cytokinesis impairment. For example, non-treated *T. brucei* presented ~80% of cells with 1 nucleus and 1 kDNA pattern (1n1k), equivalent to G1 and S phase; ~15% were 1n2k (primarily G2 phase) and 5% were 2n2k (post mitosis). Suramin treatment switched profiling and 79% of the cells accumulated in >2n, indicating the blocking of cytokinesis in *T. brucei* [121]. A similar approach can be afforded using propidium iodide followed by flow cytometry analysis. Melarsoprol-treated *T. brucei* led to the accumulation of G2/M phase from 51% to 83%, indicating increasing replication but unsegregated nuclear genome, as an evidence of mitosis inhibition [121]. Genomic plasticity is a key factor in trypanosomatids, and plays an important role that must be taken into account when developing or testing new anti-trypanosomal drugs. In this context, DNA repair mechanisms are

always being recruited, especially under stressful microenvironments like drug pressure. The enzyme uracil DNA glycosylase (UNG) participates in the DNA base excision repair (BER) pathway, and was found upregulated in *L. donovani* exposed to amphotericin B or sodium antimony gluconate. Curiously, drug-resistant clinical isolates of *L. donovani* from VL patients presented higher UNG expression [122]. suggesting that LdUNG plays a key role in BER, conferring moderate resistance to oxidants; this opens new avenues as a potential target for combination therapy against leishmaniasis. The adoption of drug discovery strategies against trypanosomatids must consider drug-resistance studies and the evolutionary role of DNA repair in this context. Antibodies can be used to track specific markers of DNA damage in eukaryotes such as the phosphorylation of threonine 130 at the C terminus of histone γH2A in *T. brucei*, which is associated with a delay in S and G2 phases of the cell cycle [123].

## **7. Genome-Wide Approaches in Target and Resistance (Resistomics)**

Functional genomics approaches are useful for identifying or validating a given drug target. This relies on strategies or tools that can be combined together with studies on drug resistance mechanisms to find clues for drug discovery. For example, the in vitro selection of drug-resistant parasites, followed by whole-genome or transcriptomic sequencing could unveil targets or signatures associated with the drug used for resistance selection. This was the case of compound 7, DDD853651/GSK3186899, selected from a chemical series of pyrazolopyrimidine scaffolds, active against *T. brucei* and used to select resistant *L. donovani* mutants as a strategy to understand the MoA and to prospect potential pathways or drug targets [99]. Whole-genome sequencing of these drug-resistant parasites revealed a single homozygous non-synonymous mutation in CRK12 (cyclin-dependent kinase 12 or cdc-2-related kinase 12), leading to a Gly 572 to Asp in the predicted catalytic domain of the enzyme, impairing electrostatic interactions and causing resistance to the pyrazolopyrimidine [99]. In this case, the resistance mechanism identification was useful to pinpoint the drug target involved in drug action. Among trypanosomatids, *T. cruzi* and *Leishmania* species (the latter belonging to the *L. (Leishmania)* subgenus) lack one or more components of the RNA interference (RNAi) machinery. However, knockdown by RNAi manipulations can be performed in *T. brucei* and *L. (Viannia)* subgenus spp., a very useful functional genomic tool to validate and identify new drug targets [124]. Inspired by these biological features, Alsford et al. (2011) described a new technique called RIT-Seq (RNAi target sequencing), where *T. brucei* were transfected with a library of interfering RNAs able to silence >99% of the mRNA in the parasite. This was followed by culturing in the presence of drug pressure in which the recovered parasites had their enriched plasmids sequenced [125]. This functional cloning technique allowed a genome-scale knockdown profiling in which the decrease of a given gene product is selected as a phenotypical marker for surviving under a stressful condition. In this way, the mechanisms underlying selective drug action and resistance can be screened in a high-throughput genome-scale RNAi panel [126]. A phenotyping genome-scale RNAi screen revealed, for example, the involvement of aquaglyceroporin 2 (AQP2) in melarsoprol and pentamidine susceptibility in African trypanosomes [127,128]. Melarsoprol is an arsenic-based drug, and similar to antimony-based compounds against *Leishmania* parasites, is taken up through aquaglyceroporin 1, which was associated with antimony resistance by using a dominant negative functional cloning strategy using a cosmid library [129]. Cosmid libraries can also be applied to select gain-of-function genes associated with a given phenotype, where the screening is based on overexpressing libraries. This approach was used to confirm previous and pinpoint new drug resistance markers in *Leishmania* parasites—a technique called Cos-seq or cosmid-based functional screening coupled to next-generation sequencing [130,131]. The most recent brother of the X-Seq family is a technique called Mut-Seq, or chemical mutagenesis coupled to next-generation sequencing. In this case, "Darwinian hands play dice" leading to stochastic mutations that could be kept when important for parasite survival under stressful pressure. This was elegantly applied to study miltefosine and paromomycin resistance mechanisms in *Leishmania* parasites. After using Mut-Seq to identify new targets and validate the essential role of kinase CDPK1 on paromomycin resistance in *Leishmania* using CRISPR-Cas9, Bhattacharya et al. (2019) suggested that

Mut-Seq screening is powerful tool to explore networks of drug resistance since CDPK1 was also involved in antimony resistance in the parasite [132]. Genome-wide approaches are very useful for capturing the main picture, and thus for choosing the most prominent biochemical pathway involved in drug action/resistance. This is also true when applying the revolutionary technique of genome editing: Clustered Regularly Interspaced Short Palindromic Repeats (CRISPR), CRISPR-associated gene 9 (Cas9)—CRISPR-Cas9. Beneke et al. (2017) developed a CRISPR-Cas9-based toolkit for the high-throughput genome editing of kinetoplastids that was further validated in single or multiple targets [133–135]. We are however currently revisiting concepts and moving from genome-wide approaches in parasite populations (or clones) to single-cell-based strategies to better understand the plasticity of *Leishmania* parasites that harbor mosaic aneuploidy—a feature that has impairments in the way the parasite will respond or not to a given drug. Using a single-cell genomic sequencing method, Negreira et al. (2020) identified 128 different karyotypes in 1560 *L. donovani* promastigotes [136]. They highlight the fact that some karyotypes presented pre-existing adaptations to antimony-based drugs, supporting a hypothesis raised even before this hint [137,138]. This reveals how complex it is to predict or open new avenues on MoA studies in trypanosomatids, and reinforces the evolutionary adaptions that guaranteed the establishment of trypanosomatids since the early Cretaceous [139]. Finally, and despite recent advances in genomic methods, there is still a relative paucity of functional annotations for a large number of gene products for trypanosomes, especially when compared to mammalian systems. In fact, this could explain why target-based methods lag behind phenotypic approaches in drug development for these parasites.

#### **8. Metabolomics in Drug Screening**

Like in a crime scene, studying the past is also a feasible alternative to tracking drug action and target identification. Metabolomics refers to the measurement of small metabolite molecules to investigate metabolic pathways, here in the context of drug discovery or target identification. Metabolite profiles are useful fingerprints offering clues on therapeutics targets in trypanosomatids, and can also be performed in the host to select signatures or markers associated with the dynamics of host–parasite interaction [140–145]. Metabolomics can also be applied to the rational development of defined minimal culture medium for in vitro drug screening purposes against trypanosomatids. In this regard, untargeted semi-quantitative or targeted quantitative metabolomics was used to decipher the major nutritional requirements of *T. brucei* and define all needs, removing unnecessary nutrients and improving drug sensitivity in activity studies [146]. Drug MoA can also be indirectly investigated through metabolomics, even without clear evidence on parasite alterations. Benznidazole is a 2-nitroimidazole prodrug that needs to be reduced in order to exert anti-trypanosomal activity against *T. cruzi*. Although benznidazole-treated parasites were minimally altered compared to untreated counterparts, metabolites concerning benznidazole linked to thiols such as trypanothione, glutathione, and cysteine indicates the thiol binding capacity of benznidazole on acting by disturbing redox homeostasis, leading to parasite death [147]. The cell redox system has also classically been related to antimony and resistance in *Leishmania* parasites. Combining untargeted metabolomics for initial screening coupled to <sup>13</sup>C traceability assays, Rojo et al. (2015) confirmed and compiled multi-target metabolic alterations not only in redox, but also in detoxification, biosynthetic processes and amino acid metabolism in *L. infantum*. Antimony-resistant parasites presented incremented proline and glutamate, supporting previous reports on high levels of glycolytic markers in resistant *Leishmania* as revealed by proteomics [148,149]. In summary, metabolomics approaches helped to identify MoA or resistance of several anti-trypanosomal drugs such as eflornithine or halogenated pyrimidines against *T. brucei*; miltefosine and antimony against *Leishmania* parasites [150]. Drug targets can also be mined in trypanosomatids by metabolomics pathway analysis using in silico approaches, as a predictive way based on pathway annotation and searching for analogous or specific enzymes [151].

## **9. Theranostic Approaches**

The term theranostic, derived from the fusion of the words therapeutic and diagnostic, is here used to define strategies designed for diagnostic purposes that also act as therapeutic agents. Dual-function molecules or smart probes can be adapted for both parasite detection/identification and anti-trypanosomatid activity. This combination of diagnosis and therapeutics is still a growing field and there are very few studies on trypanosomatids. A group headed by professors Eduardo Coelho and Luiz Ricardo Goulart in Brazil proposed the use of phage display—a high-throughput proteomic technology to generate and screen peptides and antibodies—for the serodiagnosis and prevention of leishmaniasis as a theranostic approach [152]. Using this approach, the team identified a β-tubulin from *L. infantum* that was highly antigenic and immunogenic, presenting good performance on diagnostic efficacy and eliciting Th1 response in vitro with high IFN-γ and low IL-10 levels [153]. Recently, Singh et al. (2019) reviewed the literature on nanomedicine-based approaches to circumvent leishmaniasis and concluded that much progress was made in the field reaching considerable milestones on VL nanomedicine, but translational research is needed for the coming decade for developing effective theranostic solutions [154]. Thus, many current alternatives such as liposomes, nanoemulsions, niosomes, nanodiscs, solid lipids nanoparticles, quantum dots, nanotubes, polymer conjugates, and inorganic compounds could be applied to clinical settings.

#### **10. Case Study: Proteasomal Inhibitors against** *Leishmania*

Proteasome targeted inhibitor developments by Khare et al. (2016) and Wyllie et al. (2018) are among the few major break-throughs in the quest of safe, easily deliverable, and selective drugs against trypanosomatids in recent times [155,156]. Both studies targeted the identification of a common target for intervention for *Leishmania* spp., *T. cruzi*, and *T. brucei* spp. Khare et al. began their screen with a library of 3 million compounds against the three pathogens, and identified an azabenzoxazole (GNF5343) that was effective against the three [155]. A number of substitutions leading to a less-toxic version GNF6702 further optimized the compound. In mouse model of VL and CL, with oral delivery of 10 mg kg−<sup>1</sup> for eight days, GNF6702 caused significant amelioration of liver parasitic burden. Similarly, it displayed prolific attenuation of parasite load in mouse models of Chagas disease and HAT. For leishmaniasis, Chagas disease, and HAT, the activities are comparable to the approved drugs miltefosine, benznidazole, and diminazene aceturate, respectively. In fact, for HAT it performed better than the in-use diminazene aceturate in terms of diminishing parasitic infection in brain. The primary mechanism of parasite growth inhibition by the compound series was the selective inhibition of the proteasome chymotrypsin-like activity. For analyzing resistance against the drug, they raised mutants against an early version of the drug, which showed 40-fold lower susceptibility to the drug. The phenotype was attributed to a homozygous mutation in the proteasome β4 subunit (PSMB4I29M/I29M) and a heterozygous mutation (PSMB4wt/F24L). These mutations led to reduced susceptibility to inhibition by the drug. Interestingly, the chymotrypsic catalytic center is hosted by a β5 subunit and a β4 subunit in close contact with a β5 subunit forming a plausible binding pocket for the drug. The study suggested proteasomal subunits as a selective target for the development of a common chemical scaffold against trypanosomatids. In concordance, an independent screen by Wylie et al. identified and studied a second candidate GSK3494245/DDD01305143/compound 8 [156]. The precursor of the compound was developed by scaffold hopping and substitutions from a basic component identified by a phenotypic screen of around 16,000 molecules against *T. cruzi*, and demonstrated efficacy against intra-macrophage amastigotes of *L. donovani*. The compound showed good in vitro metabolic stability (CLint = 0.8 mL min-1 g -1) and selectivity over mammalian cells. They further addressed the compound in terms of duration of treatment by rate-of-kill assay that showed that induction of cell death is achievable within 72 h at nanomolar concentration range. Pharmacokinetic profiling for bioavailability and distribution revealed that it can be orally dosed to reach efficacious levels in a range of preclinical species, including mouse, rat, and dog. Moreover, virtually no significant safety or tolerability liabilities were detected by Ames test and in mouse lymphoma cells. For identifying the mechanism of action

for the drug, the authors preliminarily adopted RIT-seq technology [125]. The study suggested that knock-down of nonessential genes of ubiquitination pathway rendered reduced sensitivity to the drug, pinpointing proteasome as the possible point of intervention for the drug. The generation of resistant mutants led to the identification of independent mutations in the β5 subunit (G197C and G197S). The mutants were cross-resistant to GNF6702. Both mutations affected proteasomal activity, as determined in vitro by UbiQ-018 label (a fluorescent label for proteasomal subunits), and the mutations resulted in insensitivity to GSK3494245 (compound 8). The proteasomal inhibitors caused cytological changes in *Leishmania* promastigotes with accumulation of vesicular structures and induced cell cycle arrest in G2/M phase. CryoEM of *L. tarentolae* proteasome in combination with compound 8 identified a number of residues from β4 and β5 subunits. Additionally, the selectivity of the drug for kinetoplastid proteasome over human proteasome could be attributed to a lack of hydrophobic interaction, as F24 in *L. tarentolae* corresponds to S23 in human and π-stacking interaction. Both works identified a suitable target for developing a common anti-trypanosomatid drug development and developed human-trial-ready molecules that precisely target chymotrypsin-like protease action of kinetoplastid proteasome without affecting the human orthologues.

### **11. Perspectives and Concluding Remarks**

At present, drugs for treating trypanosomatid diseases are far from ideal due to host toxicity, elevated cost, limited access, and increasing rates of drug resistance. Therefore, new oral, safe, short-course drugs are urgently needed. Moreover, these new drugs have to be safe and effective enough to treat patients who are asymptomatic, as well as patients who develop secondary conditions such as post-kala-azar dermal leishmaniasis [14].

In the vast majority of cases, trypanocidal agents are out of the scope of interest of the pharmaceutical industry, mainly because it is unclear how to make a profit by selling them. This situation is also becoming more frequent in the case of the discovery and development of antibiotics [157]. For this reason, drug-discovery research of novel trypanocidal compounds has been traditionally fueled by non-profit and governmental organizations. However, in the last decade, some pharmaceutical companies have become more engaged and have joined forces with academia as well as governmental and non-profit organizations to tackle NTDs. This is the case of The Drugs for Neglected Diseases initiative (DND*i*), a nonprofit research and development organization founded by Médecins sans Frontières (MSF), among other public–private partners, which has campaigned for change since 2004 to raise awareness of the trypanosomatids crisis among key policy- and decision-makers [158]. DND*i* performs high-throughput untargeted screenings of novel-drugs libraries for trypanosomatids in addition to identifying new drug candidates using targeted compounds from repurposing libraries. Since its creation, DND*i* has already provided seven treatments: ASAQ and ASMQ (two fixed-dose antimalarials), nifurtimox-eflornithine combination therapy for late-stage sleeping sickness, sodium stibogluconate and paromomycin (SSG + PM) combination therapy for VL in Africa, a set of combination therapies for VL in Asia, and a pediatric dosage form of benznidazole for Chagas disease. While combination therapies will improve the efficacy of the treatment and reduce the emergence of drug-resistant strains, currently we do not have enough effective molecules to guarantee durable therapeutic strategies. Consequently, more efforts should be deployed to discover and exploit novel families of trypanocidal drugs (with different modes of action), which could be rapidly integrated in combinatory treatments, or kept as drugs of last resort when current combinations fail.

An important bottleneck in the discovery and development of new trypanocidal drugs is the lack of well-validated molecular targets, which has traditionally hindered the use of classic target-based approaches (usually applied to the discovery of antibiotics) in the drug-discovery cascade. While it is true that this has fostered the development and implementation of sophisticated phenotypic in vitro assays, these assays encompass major challenges specific to each parasite (e.g., drugs must be active in the phagolysosome milieu when treating patients infected with *Leishmania*, drugs for HAT have to cross the blood–brain barrier, etc.). Moreover, once a hit has been identified in a phenotypic screen, different

approaches (e.g., genomics and proteomics) should be deployed to identify the specific target(s), mode-of-action of the compound, and to predict any potential mechanism of drug resistance deployed by the parasite. This information is crucial to guarantee a rational and successful optimization of the hit, and serves to develop novel target-based drug discovery cascades.

Another major challenge in drug discovery for trypanosomatids is the lack of well-defined standards/criteria (e.g., strain, culture media, incubation times, etc.) for the selection and validation of hit compounds, which sometimes leads to opposing results between different research teams. Among these criteria, one of the critical ones is the selection of the most relevant animal model that is able to mimic the pathophysiological features and immunological responses observed in human hosts (e.g., BALB/c mice vs. Syrian golden hamsters as models for *L. donovani* and *L. infantum*; acute vs. chronic models for Chagas disease, etc.).

Moreover, in order to guarantee the success of drug discovery/repositioning in the fight against trypanosomatids, we have to generate high-quality data in many endemic countries (including field strains, drug-resistant strains, etc.), and to do so, we have to effectively increase the engagement of endemic countries in the R&D process [159].

New powerful and robust in vitro, in vivo, and in silico technologies have emerged in the last ten years. Moreover, we now have a more refined knowledge of the biology of these parasites, as well as the unprecedented ability to surgically manipulate trypanosomatids genome. The optimal use of these tools and knowledge will undoubtedly accelerate current drug discovery cascades, leading to the delivery of satisfactory treatment options for neglected patients with trypanosomatid infections.

**Author Contributions:** Conceptualization, C.F.-P., A.B. and R.L.d.M.-N.; writing—original draft preparation, C.F.-P., A.B., R.L.d.M.-N. and A.C.; writing—review and editing, C.F.-P., A.B. and R.L.d.M.-N.; project administration, C.F.-P.; funding acquisition, C.F.-P. All authors have read and agreed to the published version of the manuscript.

**Funding:** This work was supported by a Natural Sciences and Engineering Research Council of Canada Discovery Grant RGPIN-2017-04480 and by the Canada foundation for Innovation (www.innovation.ca), grant number 37324; both awarded to CFP. AC is supported by Fonds de Recherche du Québec-Nature et Technologies (FRQNT) scholarship program. RMN is a CNPq (Brazilian National Council for Scientific and Technological Development) Research Fellow (#310640/2017-2).

**Acknowledgments:** The authors want to thank Aida Mínguez-Menéndez for her help with the conception of the figures.

**Conflicts of Interest:** The authors declare no conflicts of interest.
